SEATTLE – Applying single cell transcriptomics to brain tissue from living donors is laying bare the vast differences between the human brain and the mouse brain, calling into question the use of animal models in psychiatric drug discovery.
While the research indicates the majority of basic human cell types have a mouse equivalent that has been conserved over 75 million years of evolution, it highlights extensive differences between homologous human and mouse cell types. Functional analyses of living human brain cells show there are many differences in how and when genes are expressed.
As one notable example, serotonin receptors that are key to appetite, mood, memory and sleep and other brain functions – and the target of a number of drugs – are found in different kinds of neurons in the two species.
“This is really salient to drug discovery,” said Ed Lein, director of the Human Cell Types Program at the Allen Institute for Brain Disease. “The paradox is, the mouse is deceptively similar, but the details vary significantly,” he told attendees of the American Association for the Advancement of Science meeting.
Lein and colleagues previously used postmortem tissue to map the organization of individual cell types in different layers of the human cortex. “This shows the brain is extremely cellularly complex, but at the same time finite and definable,” he said.
That work, carried out as part of the NIH’s Brain Initiative, now forms the framework to investigate cellular function. “Once we know what the cells are, we can ask what their properties are,” Lein said. “This calls for a new paradigm; we need to look in living tissue.”
But that presents a huge problem. “The human brain is hugely inaccessible,” Lein said. To address that, over the past few years he has established a network of Seattle neurosurgeons through whom he gets access to tissue samples from patients having resections to treat intractable epilepsy or brain tumors.
He now receives 50 consented donations per year. “One surprise is how remarkably healthy the tissues are,” Lein said. Mouse brain tissue remains alive for six hours, whereas it is possible to keep human brain tissue alive for six to seven days, and to culture cells for longer. “This was a surprise – we don’t understand the mechanisms behind it,” said Lein.
The electrophysiology of living human brain cells is tested using the traditional patch clamp technique; they are injected with dyes to reveal their morphology; finally the transcriptome of the nucleus is analyzed to profile the gene expression.
“Transcriptomics [data] is highly predictive for human cortical neuron properties,” Lein said. You can also compare across species to see if properties are conserved. There are some similarities, but many features are different.”
There also is more variety between different types of human neurons than there is in mouse types.
With the detailed human brain cell profiles, Lein and colleagues are establishing what he calls a “periodic table” of neuron types and their properties. It is not known how many different types of cells there are, but so far 100 have been identified, most of which have never been profiled before in such detail. “The brain is complex. I anticipate there will be several thousands of types of cells in the brain,” Lein said.
Given those insights, it is not surprising that the rate of success in using animal models in preclinical research has been low. However, Lein is not suggesting that mice and other models, such as brain organoids or brain cells derived from induced pluripotent stem cells, have no value in neurological research.
Based on research in living brain cells, it will be possible to make detailed comparisons of mouse and human brain cell types and to validate decades of research in rodent models. It also is possible to see whether the target of a drug is expressed in the same way in human and mouse brain cells.
“Model organisms have enormous utility, but it is important to understand what they model. Not even organoids or stem cells are as good as living tissues,” Lein said.
He suggests neuroscience has reached the stage the field of immunology was at several decades ago. It is now well known that the immune system runs on a network of many different cell types, and Lein said he hopes to reach that level of cellular resolution in neuroscience, reducing the complexity of the brain to its component cell types, defined based on their function, anatomy, connectivity and gene expression.